| Earthquake is a very fearful natural disaster that occurs suddenly and endangers people's life and property. A destructive earthquake may sometimes causes a lot of property losses and casualty in a very short time period less than one minute. The government, scientists and public all want to know the damage area and severity caused by an earthquake as soon as possible. The information is also the important reference to the effective organization for rescue and reconstruction.Remote sensing image can be acquired very fast, can mirror the objective world entirely and intuitionisticaly, and provides a new information source for earthquake damage rapid evaluation. It is convenient, fast, effective way to evaluate earthquake damage with remote sensing data, is all along a problem commonly attended by researchers in both areas of disaster prevention and reduction and of remote sensing technology application. Went through the initial stage of visual estimation, remote sensing image digitally processing based earthquake damage rapid evaluation method gets a delightful progress, along with the very fast development of digital image process technology, and the enhancement of positioning accuracy of remote sensing spatial data. The earthquake damage evaluation method by means of remote sensing image change detection and digital features that recently published, is still in research stage, one of the bottlenecks to update is that there is a lack of validation by enough earthquake cases. It can be expected that some new problems must be found and some improvements must be made in a lot of validation of case studies. The remote sensing images pre- and post Bachu earthquake on February 24, 2003, in Xingjiang acquired by satellite SPOT, are collected in this paper. A change detection and the image texture feature analysis are carried out. A probabilistic model for earthquake damage evaluation is preliminarily built by comparing the results with the detail information from the post earthquake ground survey. A validation of one earthquake case study is worked out.Firstly, the basic data on Bachu earthquake on February 24, 2003, inSinkiang is summarized, the spatial distribution of the earthquake damage from the post earthquake ground survey is emphasized, the remote sensing images pre- and post event acquired by satellite SPOT and other related basic data are introduced.Secondly, spatial matching, the key point in spatial analysis of remote sensing image data, is emphasized and described in detail, the elements of geometric correction on remote sensing image by multinomial method and the technical points on selection of ground control points are summarized.Thirdly, after the elements and objects of grey scale matching of images is analyzed, an idea to improve the grey scale matching of pre-and post earthquake remote sensing images is presented, in which.the areas outside of residential areas are processed as objective areas of matching. The image pairs for change detection are prepared, the grey scale difference and ratio are calculated by means of ENVI software. The statistic analysis for the residential areas with post earthquake ground survey data is carried out, the values of mean and standard deviation of regions with intensities IX, VIII and VII are obtained, then a preliminary probabilistic model for earthquake damage evaluation is built. As internal check, the probabilities of three intensities in every residential area are calculated from the values of mean and standard deviation of the grey scale in the area, the intensity with the maximum probability is compared with intensity in the area from post-earthtfuake ground survey, to show the reliability of the method developed in this paper.Finally, the analysis method of texture statistic based on neighborhood is be used to extract the texture feature of residential areas both in the 10m and 2.5m ground resolution image of SPOT5 satellite after quake. Counting the mean and...
|
PDF Full Text Request